Collama is a minimal and practical setup to run Ollama models directly on Google Colab.
It is designed for developers who don’t have a local GPU or want to quickly test LLMs without complex setup.
- One-click Ollama setup on Colab
- Run models like Llama, Qwen, Deepseek, CodeLlama
- LangChain integration (prompt → response pipeline)
- Minimal setup, maximum speed
- Developers without a local GPU
- People who want fast prototyping
- Anyone experimenting with LLMs
- Builders working on agents or automation
- Test coding LLMs
- Experiment with prompt engineering
- Build simple AI tools
- Create LangChain + Ollama workflows
https://colab.research.google.com/github/0x1881/collama/blob/main/Ollama_Setup.ipynb
Requirements:
- A Cloudflare account
- A domain connected to Cloudflare
Steps:
- Go to the Cloudflare Zero Trust panel: https://one.dash.cloudflare.com/
- Navigate to Networks > Connectors > Create a tunnel > Cloudflared
- Give your tunnel a name
- Select your device’s operating system (Debian)
- At the bottom, copy the token from the command:
- cloudflared tunnel run --token ***
- Add this token to Colab Secrets with the name CLOUDFLARE_TOKEN
- Colab sessions are ephemeral (restart resets everything)
- GPU usage is limited
- Large models may not fit in memory
This repository is open to contributions.
If your notebook uses Ollama to solve a real use-case that can help others learn something, feel free to open a pull request.
This repository is not intended for production use.
It is designed as a quick test and playground environment.
This project was inspired by https://github.com/5aharsh/collama.
Thanks to the author for the initial work and concept.
